Though now less common in developed countries than previously, in 2003 cervical cancer is projected to result in 12,200 new cases and 4,100 deaths in the United States. These relatively low rates are attributed to the success of cervical cancer screening and treatment of cervical intraepithelial neoplasia (CIN). To reduce further the incidence and mortality of cervical cancer in developed countries while avoiding unnecessary procedures and conserving resources is a challenge that confronts clinicians and health policy makers. The recent move away from recommending lifelong annual cervical cancer screening towards consistent screening every two to three years is one example of an effort to meet this challenge. Recent estimates for the US indicate that more than one million women are diagnosed with low-grade neoplasia (CIN I) annually, and that about 500,000 women will have higher-grade lesions (CIN 2 or CIN 3). Follow-up strategies must strike a balance between finding and treating persistent or recurrent lesions and potential overuse of procedures and resources. The recent publication of the findings from the ALTS trial has led to recommendations for the management and follow-up of CIN. These recommendations include follow-up after treatment at 4 to 6 month intervals until three negative cytology results are obtained, followed by annual screening. The duration of annual screening remains unspecified, however, and the long-term risk and time patterns of recurrence of CIN or invasive cancer after treatment remain unclear. The relative costeffectiveness of these recommendations has not been evaluated. Given the large number of women diagnosed with CIN every year, over time a substantial number of women will be assigned to long-term annual cytology for post-treatment surveillance. Our study will examine the long-term risks of recurrent CIN and changes in risk over time in a cohort drawn from a comprehensive, population-based dataset. Based on this data and previous work, we will conduct a cost-effectiveness analysis to compare strategies for longterm follow-up of women who have undergone evaluation and treatment of CIN. Data from the cohort study and a systematic review of the literature will be used for recurrence risks, costs will be estimated by coding of clinical pathways, and utilities obtained from diverse populations will be applied to a previously validated Markov model. Alternative strategies for post-treatment surveillance will be evaluated. [unreadable] [unreadable] [unreadable]